THE ONTOLOGY OF SPECIES:

A RADICALLY PLURALISTIC PERSPECTIVE

By

BRANDON D. HOLTER

A thesis submitted in partial fulfillment of the requirements for the degree of

MASTER OF ARTS IN PHILOSOPHY

WASHINGTON STATE UNIVERSITY Department of Philosophy

AUGUST 2009

ii

To the Faculty of Washington State University:

The members of the committee appointed to examine the thesis of BRANDON

DAVID HOLTER find it satisfactory and recommend that it be accepted.

_________________________________________________

Joseph Campbell (Chair)

________________________________________________

Matthew Slater

________________________________________________

Michael O’Rourke

iii

THE ONTOLOGY OF SPECIES:

A RADICALLY PLURALISTIC PERSPECTIVE

Abstract

By Brandon D. Holter, M.A. Washington State University

August 2009

Chair: Joseph Keim Campbell Species pluralism is an increasingly popular position in philosophy of biology,

and with good reason. The diversity of the biological world and the evolutionary forces

that shape them has convinced many biologists and philosophers that there are many

overlapping groups of organisms we might call species. It is typical to approach the

species problem with the question “what are species?” but I suggest the question “What

aren’t species?” is equally daunting. Specifically, what sets of organisms do not

constitute species? I content that no metaphysical distinction can be drawn between

species groups and random assortments of organisms.

Species have posed conceptual problems [for philosophers] for centuries. There

are many competing theories and little consensus. In chapter two I review the standard

problems for species monism and the arguments for pluralism. I argue that not only are

many species theories equally legitimate, but that the plurality, interactivity, and dynamic

nature of evolutionary causes necessitates a pluralistic species theory.

iv

Moderate species pluralism must justify the exclusion of species theories on the

basis of some principle of moderation. These principles specify why some species taxa

are real and others are not. Without such a principle, pluralists will have to admit that

“anything goes” with regards to the ontology of species. In the penultimate chapter, I

examine the parallels between mereological universalism and species pluralism.

The important congruency between the species debate and mereology is the

problem of moderation, its motivations, intractability, and possible resolution. While the

problem of moderation is present in both domains, I argue that philosophers of biology

have failed to take it as seriously as mereologists. As a result, the solution offered in

mereology is instructive for species pluralists.

v

TABLE OF CONTENTS

Page

ABSTRACT………….......................................................................................................iii

CHAPTER

1. INTRODUCTION…...........................................................................................1

THE “DISORDER” OF EVOLUTIONARY BIOLOGY............................6

OVERVIEW…............................................................................................9

2. AGAINST MONSIM AND FOR PLURALISM ABOUT SPECIES………...15

TYPOLOGICAL MONISM……………………………………………..18

HISTORICAL AND INDIVIDUALISTIC MONSIM…………………..22

FOR PLURALISM………………………………………………………26

3. THE PROBLEM OF MODERATION………………………………………..34

PRAGMATIC PLURALISM……………………………………...…….38

THE EPISTEMOLOGICAL APPEAL TO SCIENCE...………………...43

THE CAUSAL APPEAL TO SCIENCE………………………..……….46

4. SPECIES UNIVERSALISM AND MEREOLOGY……...…………………..54

REALISM…………………………………………………………...…...61

QUANTIFIER RESTRICTION AND VAGUENESS………………..…63

5. CONCLUSION………………………………………………………………..68

BIBLIOGRAPHY………………………………………………………………..72

1

CHAPTER ONE

INTRODUCTION

Well-formed answers to the question “what are species?” tell us by virtue of what

relationship numerous organisms come to comprise a further ontological entity called

species. The species question is best construed as the question “What relation(s) must

some x’s (organisms) instantiate to form a species?” Van Inwagen’s special composition

question, which informs this formulation of the species question, asks what relations must

hold between parts for them to compose a further whole object, but other ontological

inquires are well served by the general form of the question as well (van Inwagen 1990,

21-2; Slater ms. 6). I call individual relations proposed as answers to the species question

species relations.

Each species relation corresponds to a species concept, a theory about what

species are that is defined by the relation that underwrites it. The “biological species

concept” (BSC) identifies reproductive compatibility as the relation that identifies all

species taxa (Mayr 1992, 17). Reproductive compatibility may be interpreted variously,

but it roughly means the ability to procreate. Any species theory (taxonomic approach)

identifies which relations are properly considered species relations. The difference

between a species theory and a species concept is that theories may identify multiple

species relations, hence many corresponding concepts. Species concepts are necessarily

monistic. Species taxa are real species, whatever they turn out to be, while species

theories (taxonomic approaches) are merely proposed answers to the species question.

2

This work attempts to establish the intractable difficulties dormant in various

attempted answers to the species question without denying the reality of species. Species

do, after all, appear discreet to the casual observer. One notices with little effort that some

organisms appear extremely similar; often members of species are indistinguishable to

the untrained human eye, while those members appear distinctly different from other

groups of similar organisms. “No naturalist would question the reality of the species he

may find in his garden, whether it is a catbird, chickadee, robin, or starling… Species at a

given locality are almost invariably separated from each other by a distinct gap” (Mayr

1987, 146).

I examine issues of realism, anti-realism, monism, and pluralism in the species

debate. These alternative positions lead to a natural division of species theories into four

categories, monistic realism, pluralistic realism, monistic anti-realism, and pluralistic

anti-realism. The distinction between monistic and pluralistic anti-realism is not

metaphysical as both positions contend that there are no metaphysically real species taxa.

The monism/pluralism debate for anti-realists depends practical relevance alone

(Stanford 1995, 89). They argue that the term ‘species’ ought to be used certain ways, not

that certain species theories denote real taxa. Arguments from anti-realists regarding how

best to conceptualized species take place on the semantic or pragmatic level, not at the

level of metaphysics.

Some answers to the species question are monistic. They insist that only one

relationship between individuals is pertinent to species membership and that each and

every species will instantiate this relation. The aforementioned biological species concept

offers one such theory of species monism. If a monistic species theory is tenable, no set

3

of organisms lacking the favored relation may rightly be considered a species and any

organisms displaying that relation must be members of a common species. Thus

opponents of monism often point to problematic cases of species membership in order to

undermine particular monistic species theories.

Problematic cases come in two varieties: unrecognized taxa and illegitimate taxa.

Groups of organisms that do not display the relation endorsed by a specific theory but

nonetheless ought to be considered species are instances of unrecognized taxa. Infertile

organisms that are highly physiologically and genetically similar pose an unrecognized

taxa problem for the BSC. Illegitimate taxa display the favored relation without forming a

group that we are likely to consider a species. When members of two distinct species, Sx

and Sy, produce fertile offspring, the reproductive ability of Sx and Sy should not suffice

to group them under a single conjoined species. Every instance of hybridization between

distinct species presents the BSC theorist with an illegitimate taxa problem because the

relation of reproductive compatibility is present between Sx and Sy despite there being

no common species taxa (Sterelny and Griffiths 1999, 188; Stanford 1995, 73).

Both kinds of problematic cases compare the taxa identified by a target species

concept, in these cases the BSC, and assert that they fail to live up to intuitions about

species taxa. One might object that these modes of criticism wrongly assume that

alternative species relations are legitimate, such as the genetic and physiological relations

identified in the unrecognized infertile taxa example. But mere intuitions are not the only

basis for these objections, they are informed by the biological facts, and these are the

only facts we can appeal to in critically analyzing species theories.

4

Species pluralism asserts that there is more than one relation organisms of a

common species may bear. Some sets of organisms form species because they instantiate

one species relation, others are individuated by another relation. Perhaps asexual species,

being incapable of the reproductive compatibility needed for BSC, are distinguished by

their genetic make-up. Such a theory recognizes the taxa formed by the relation of

reproductive compatibility as well as the relation of similar genetic constitution. For

pluralists, all sets of organisms bearing any of the relations they endorse as species

relations count as real species taxa. Which relations the pluralist adopts will determine

which taxa count, so the substance of any pluralistic theory lies in why some relations

and their corresponding species concepts are considered real taxa.

Pluralism faces a problem not shared by monistic theories of species.

Monists claim that one relation between organisms defines species while pluralists

conjoin multiple relations. One question that plagues pluralism is why only those

relations should count as definitive of species (Stanford 1995, 77). Monists, by assuming

that taxonomists must strive for a univocal account of species, are able to exclude

alternative theories by default (Sober 1984, 341). Only one relation defines species, so

once a monistic theory is taken to be an adequate account, all other species concepts must

be illegitimate. By denying the uniqueness of species relations, pluralists are made

susceptible to the pointed question “What makes these relations adequate grounds for

species membership and other relations illegitimate?”

Pluralists need a principled stance on what makes their favored relations

distinguishable from those they reject, a principle of moderation. Moderation principles

answer the question “What makes the taxa instantiating relations R1 R2 through Rn,

5

different from other relations such that R1 through Rn constitute species relations while

the others do not?” There must be something common to the relations that distinguish

species taxa lacking in other non-species relations. Perhaps only scientific species

concepts, and their corresponding relations, should count. Rather than answering the

species question by identifying a unique relation held by all species, pluralists identify a

principle for identifying all of and only those relations. Thus principles of moderation

define pluralistic species theories by picking out species relations, whereas identification

of a single relation defines monistic accounts.

Every species theory identifies what commonality all species taxa share. Monism

holds that all species members display a particular relation. A principle of moderation

indicates what unifies a plurality of species relations under the heading species taxa. A

similar question may be posed with regard to the species rank or category. The species

rank question asks what makes species taxa different from higher Linnaean categories,

like family and genera, as well as lower order taxa (subspecies). Questions about species

taxa should remain distinct from issues about the species category (Devitt 2009, 2;

Ereshefsky 1992). I am particularly concerned with what makes some groups species

taxa, not with what makes those taxa belong to a particular rank.

The species taxon question involves why particular organisms belong to a species

taxa, while the species rank question asks whether particular taxa are rightly considered

species. Moderation principles do answer the question “what do all species taxa have in

common?” but they do not necessarily entail any position with regards to the what

distinguishes the Linnaean category of species from other such categories. A pluralist

may hold that there is nothing distinguishing species taxa from other Linnaean categories,

6

even that there are no Linnaean categories at all, while at the same time contending that

something distinguishes species from random collections of organisms (Ereshefsky

2001). Some pluralists contend that term ‘species’ has no univocal meaning, so there can

be no category to which each species taxa belongs (Ereshefsky, 1992). Confusing

questions about taxa with rank or categorical questions unnecessarily muddies the waters

of species theorizing.

The “Disorder” of Evolutionary Biology

Some are attracted to the study of natural science because they see a beautiful

harmony in the orderly structure of natural phenomena. This examination of the

metaphysics of species takes as fundamental to evolutionary biology the diversity and

disorder of evolutionary phenomena. The contrast between chemical elements and

species illustrates the difference between an orderly or unified picture of the natural

world and the diversity found in the biological sciences. In the elements found on the

periodic table, the microstructure of the substance strictly determines both what element

it is and what visible properties it exemplifies (Sober 1980, 332).

Gold is always constituted by the same microstructure and has the same number

of protons, and instantiates observable properties by virtue of this microstructure. Thus

the weight and electrical conductivity of gold, for example, are caused by its

microstructural properties and each of these is found in every example of gold as well. If

we found a sample of gold differing in observable properties, say pink in color, scientists

would be shocked. Not so in biology. “The patterns of variation actually found in nature

do not fit easily with the idea of an essentialist definition of species” (Okasha 2002, 196).

7

Discovering organisms with some novel feature is not surprising in the slightest. In fact,

it is what biologists have come to expect because the nature of the evolutionary process is

such that variation is the rule rather than the exception in biology.

The theme that drives my arguments is the variability of the products and

processes of evolution. The process of speciation is a slow and gradual one measured in

generations of organisms with intermediate individuals along the way (Dupré 1981, 90).

There are, on the other hand, no intermediate steps between two elements on the periodic

table. The nature of evolution is not to produce clean-cut divisions the way the universe

seems to have produced clear distinctions in elements1. David Hull observes that “our

inability to distinguish most species by sets of necessary and sufficient conditions follows

from evolutionary theory just as surely as quantum indeterminacy follows from quantum

theory” (Hull 1976, 180).

Evolution, the force that produces variation and similarity in organisms, is itself a

diverse set of processes. These processes produce individual organisms distinguishable in

some ways from most other organisms, but relatively similar in others. Thus we notice

groups of individual organisms similar to each other and dissimilar to other such

collections—species. Much of the disagreement over species has been the result of

different perspectives on how best to reflect the role of evolutionary processes in a theory

of species membership. The claim that species taxa should reflect the processes that gave

rise to them is not exclusive to evolutionary theory, however. Creationists contend that

species are determined by god’s creative will. Legitimacy of “creation science” aside,

1 Sober (1980, 334-5) thinks chemistry has no such clean-cut boundaries and that this is no problem for essentialism. He argues that transmutation, conversion of one chemical element into another, provides a case of vague boundaries. But transmutation is hardly the natural state of elemental matter.

8

even before the theory of evolution came along it was recognized that species taxa might

be determined by the causes that generate them.

Evolutionary biologists recognize that there are many causes that contribute to the

formation of species. Darwin understood that both environmental pressures and sexual

reproduction were key processes in his theory, but in addition “many processes Darwin

never dreamed of are important in molding populations, including mutation, segregation,

recombination, genetic drift, gene conversion, and meiotic drive” (Richerson and Boyd

2005, 5). That there are multiple causes of variation and similarity between individuals is

not itself problematic for species theories. However, because evolutionary causes interact

with one another and occur with different prominence in particular speciation events, it is

difficult to derive species taxa from the operation of evolutionary causes (Chandler and

Gromko 1989).

The causal mechanisms that produce species taxa are not only many, but they do

not act independent of each other. Evolutionary causes are interactive. When some new

trait arises among a population it will have both a genetic and non-genetic factors that

figure into its evolution (Sterelny and Giffiths 1999, 98). The genetic causes of a trait are

not independent of environment in which it occurs; genes may be affected by other

genetic factors, its genetic environment, as well as the ecological environment in which

the organism finds itself. “Taken together, the relations of genes, organisms, and

environments are reciprocal relations in which all three elements are both causes and

effects” (Lewontin 2000, 100).

Evolutionary causes produce disorderly species taxa because they are many,

interactive, and dynamic—they occur with variable importance from case to case.

9

Sometimes environmental pressures will be highly influential, sometimes not so much.

This accounts for some of the taxonomic differences between chemical and biological

kinds—their causes are dissimilar. I have spoken of evolutionary mechanisms as if they

strait-forwardly caused species. This is, of course, not the case, but there is a sense in

which such talk is justified. While tigers are not, in a strict sense, caused at all, their

distinguishing features are. Species are differentiated by the properties unique to them

and it is evolutionary mechanisms that produce these unique properties. Insomuch as new

species are produced in evolution, they are produced by the changes in properties that

occur via evolutionary forces. In this sense, when a new species becomes noticeably

distinct from its parents species, we can say that the species was “caused by” the

evolutionary forces that produced its unique features.

Overview

Motivated by the plurality of dynamic and interactive evolutionary causes, I

present a theory of species that is radically pluralistic. More sets of organisms are really

species taxa than most laypersons or scientists believe. The arguments that produce the

particular theory of species pluralism I offer rely on a process of elimination of sorts.

Chapter II reviews the problems for species monism and the standard arguments for

pluralism. First, I review negative arguments against some prominent theories of

typological monism based on the inadequacy of the species relation that defines them.

Every species theory has a criterion for what should count as a real species and each

theory is criticized on the grounds that their respective criteria are not tenable forms of

species realism. I content that historical and individualistic accounts of species suffer

10

many of the same difficulties as well. After examining the problems faced by monism,

two motivations for pluralism are presented.

The first positive argument for species pluralism holds that multiple species

theories are legitimate, so the assumption of monism is wrongheaded. To the extent that

any one species theory identifies real taxa, other theories seem to capture real taxa just as

well. A second motivation for pluralism stems from the observation that discontinuity,

change, and variation are the rule rather than the exception in biology, a situation I have

been painting in broad strokes so far. Many disputes about species revolve around which

property ought to be considered primary in our taxonomic practice. Which properties of

organisms—physical appearances, genetic constitution, historical relations, ecological

niche, etc.—ought to determine species membership?

Some theories of species, most notably natural kind theories, have contended that

what is important about species is that they have many properties in common. Species

kinds on this view are, minimally, property clusters because sharing a number of features

in common with other members is simply definitive of species (Boyd 1999, 406-7). Many

pluralists, on the other hand, stress that the properties biologists find most salient for

species are distributed in ways that do not correspond or co-occur in all of the members

(Mishler and Donoghue 1982, 494). Finding any one property P in a population of

individuals neither guarantees that P will be found in all of the individual members of the

species, nor that any other properties will be clustered together in individuals displaying

P. For pluralists of this bent, no property or trait is guaranteed to occur in individuals

based on the presence of some other property. All to often, correspondences between

11

salient properties have simply been assumed by biologists and philosophers alike (Ibid.

494; Kitcher 1984, 330).

We might analyze the distribution of any property across the entire population of

the living world and choose to organize taxa based on its presence or absence. However,

if we map the distribution of many biologically important properties among organisms,

we will find that each property is unique in its distribution. As a result, each approach to

classification that corresponds to the recognition of certain properties will model a

distinct set of species taxa and distinct patterns of biological diversity. The essence of this

motivation to pluralism is that diversity among individuals and species is so prevalent

that no univocal account of species is possible, for it will fail to capture all of the relevant

aspects of species taxa. Disorder and variety trickle down from evolutionary

mechanisms—to species as well as individual organisms.

Chapter III focuses on pluralistic species theories. Monism faces the challenge of

identifying one criteria of species membership that satisfies every occurrence of every

species taxa. Pluralism faces a higher order issue. Monism has the advantage of ruling out

every other species theory available based on the contention that only one theory can be

legitimate. Pluralism must justify the exclusion of species theories on the basis of some

principle of moderation that specifies why some species taxa are real and others are not.

Otherwise the pluralist will have to admit that “anything goes” when it comes to

identifying and defining species taxa. Few species pluralists have devoted much energy

to defending against this problem of moderation, but it is a pressing concern (Stanford

1995, 77; Ghiselin 1987, 135-6)

12

In the penultimate chapter, I examine the parallels between mereological

pluralism and species pluralism. Mereology, the theory of parts and wholes, is an

especially applicable subject given the recent enthusiasm for treating species as

individuals rather than kinds (Hull 1976; Ghiselin 1975; 1987). I concur with Phillip

Kitcher that this distinction in ontological models is biologically neutral; all species

theories can in principle be construed as either individualist or kind-based (Kitcher 1984,

314; 1987, 187). The debate between the two metaphysical perspectives is trivial for the

taxonomic purposes of biology.2

The popularity of arguing that species are individuals or essentially historical

entities is a relatively recent development in the philosophy of biology. These theories

emphasize that species should be historically connected lineages, not just groups of

organisms sharing common properties (Griffiths 1999; Ghiselin 1987). The latter view

relies on classes, types, or kinds as the organizing principle of species. Individualists

about species contend that the organisms should be grouped according to the part-whole

relationship rather than kind and membership. Though not all proponents of the historical

conception endorse the species as individuals thesis, Griffiths most especially does not,

there is little that distinguishes individualism from mere historical accounts.

Both historical and individualist accounts of species are inspired on the principle

most prominently promoted by Ernst Mayr that species are cohesive units distinguished

by the individuals’ reproductive relations (Mayr 1992, 17). The picture of species taxa

2 Matthew Slater has argued that mereological species differ from kinds because kinds may be amenable to vagueness, whereas there can be no such thing as a vague object. I reject any account of species as vague for two reasons: 1) It is an ad hoc way of avoiding problematic cases. Any unrecognized or illegitimate taxa can be waived away with the claim that “it’s just vague.” 2) Lewis’ invective “What is this thing such that it sort of is so, and sort of isn’t, that there is any such thing?” seems applicable whether the thing be kind or individual (Lewis 1986, 213).

13

that emerges is one of organisms sharing certain relations, sexual relations that require

spatio-temporal contact, rather than sharing similarities or identities as on typological

accounts. Thus individualists and historical species theories view species as spatio-

temporally connected entities. The idea that these relations are anything like typical part-

whole relations is controversial, I believe unfounded, and most importantly, trivial for our

purposes. An imaginative philosopher can understand the idea that species might be

groups of historically connected and reproduction-relationally defined individuals

without reference to mereology.

The important congruency between species and mereology is the problem of

moderation, its motivations, and its possible resolution. The relation that parts must have

to be properly considered a whole object—the parthood relation—is subject to rigorous

debate (Markosian 1998, Van Inwagen 1990). As in the species debate, there is a variety

monistic, pluralistic, realistic, and anti-realistic perspectives, with the pluralistic realists

in need of a principle of moderation. While the problem of moderating pluralism is

present in both mereological and species theories, I argue that mereologists have taken it

more seriously and offered a more satisfactory response than philosophers of biology. As

a result, the solution of the problem of moderation in mereology is instructive for species

pluralists.

14

CHAPTER TWO

Against Monism and For Pluralism About Species

Many of the arguments presented by species pluralists are variations on three

general themes. One strategy targets a prominent monistic account of species and shows

that there are problematic cases. Either the theory does not account for every case of

species, there are unrecognized taxa, or some species endorsed by the theory are

illegitimate. A substantial number of counter examples to one favored theory of species

monism, a species concept, might succeed in refuting that theory but it quite obviously

does not directly support pluralism. One problem that pluralists continually encounter

when trying to advance these kind of negative arguments is the obstinate insistence that

some other revised theory of species will eventually be discovered that will apply to

every species taxa (Dupré 1996, 443).

The possibility always remains that the fault is not in the nature of our concepts, but in our lack of insight into their determinant structure... a few more centuries of frustration (at least) would be needed before we can declare the project hopeless. For the foreseeable future, pluralism is the ‘null hypothesis’ that we should attempt to refute (Sober 1984, 340-41). While arguments against particular species theories do not themselves make good

arguments for pluralism, it should be noted that for those who assume species monism is

tenable, showing the inadequacy of a commonly held or highly regarded monistic theory

is persuasive in ways that other modes of argumentation are not. They force the

committed monist to form a counter argument in defense of the theory, revise the theory,

or find some new theory to adopt. The argument that eventually some monistic account

15

will prove satisfactory merely asserts the validity monistic intuitions in the face of

substantial challenges. In the second section of this chapter, I review the flaws found in

monistic accounts of species.

To contrast, positive arguments that motivate species pluralism are susceptible to

responses to the effect of “that’s an interesting theory, but I prefer my own monistic

account.” The most common way of promoting pluralism is to argue that there are many

legitimate ways of dividing up species taxa. Different philosophers rely on different

interpretations of legitimacy, however, and this marks the distinction between variations

on the equal legitimacy strategy. I consider three such variations. The utility of a species

concept may ground its legitimacy by identifying features of organisms that are useful for

some purpose. On this view, taxa regarded by the scientific community as mere folk

taxonomy concocted for specialized purposes are legitimized, considered real, just

because they are practical.

John Dupré argues that such common sense or “ordinary language” taxa are

legitimate because they are useful and our ontology would be remiss if it did not include

them (Dupré 1993, 28-36). A more conservative equal legitimacy argument is that

different scientific practices require different species concepts. For instance, even if

species are best defined in terms of their lineage of decent (as per historical species

theories) paleontologists will never be able to identify a Tyrannosaurus Rex fossil as part

of that species by investigating its reproductive relations. Whatever criterion they do use,

likely gross physical properties of the fossil, can only be said to be legitimate if those

properties reliably indicate species taxa. Thus the legitimacy of a scientific practice might

buttress the legitimacy of a species theory that utilizes it (Kitcher 1984, 326). The third

16

equal legitimacy argument appeals to the relevance of linking species concepts with

species generating processes3. Many species theories are considered legitimate in part

because they highlight causal processes key to evolutionary theory, the aforementioned

criteria of reproductive relations included. Based on this, Marc Ereshefsky argues that a

plurality of species concepts reflect relevant causal process and so should be counted

equally legitimate candidates for species taxa (Ereshefsky 1992, 350-356).

Lastly, I consider a third strategy for motivating pluralism that turns on that idea

that once we reach a proper understanding of the biological and evolutionary facts, the

expectation that species monism should obtain is rendered obsolete. Such arguments try

to broadly characterize the evolutionary process and its products in order to show that

there is little reason to expect that there is a single taxonomic approach that identifies all

and only the species taxa. Once we recognized the diverse, interact, and dynamic nature

of evolutionary causes, and that the organisms produced are highly variable in ways that

chemical elements like gold are not, species pluralism becomes an expected outcome of

the evolutionary process. On one hand, the variability of individual organisms undercuts

monistic theories because, as the following section shows, no common properties or

relations among organisms can identify every species. On the other, the multiplicity and

interactivity of evolutionary causes prevents biological taxonomy from developing a

monistic theory that captures a single causal process responsible for every species

produced. There is, therefore, no reason to expect that species monism should hold.

Taken individually, negative arguments against monistic theories, positive

arguments for equal legitimacy, and the argument from evolutionary disorder are not

3 Ray Chandler and Mark Gromko (1989) oppose such a linking.

17

fully compelling in the sense of obligating any rational person to accept the conclusion of

pluralism. Indeed it is not clear what a proof of species pluralism would look like.

However, as pluralists have developed these separate lines of argumentation, their

conjunction has become a powerful case against monism. If no single candidate theory is

without substantial problematic cases and multiple theories seem equally adequate for

delineating some species taxa, then this should be enough to at least question the utility of

holding onto monist intuitions. Of course, nothing dictates that there could not be a

monistic theory waiting to be discovered even if this is the case. It may be impossible to

prove beyond any possibility of doubt that there is a plurality of overlapping species taxa,

but the deficiency of monism, the legitimacy of pluralism, and a proper understanding of

evolutionary facts do make a formidable case for the thesis.

Typological Monism

The traditional model for kinds or types4 of objects — the model which species

were assumed to conform to since the time of Aristotle — distinguishes kinds by their

intrinsic properties (Hacking 1991, 120-21). Essential properties, the intrinsic properties

that serve to identify and define classes, are thought to be causally responsible for at least

some of the observable features of individuals of that kind (Kornblith 1993, 35). On this

view, all members of any particular species, like chemical kinds such as gold and water,

should turn out to have some intrinsic property in common. Unfortunately for species

4 I use these terms interchangeably.

18

monists committed to the traditional account of kinds, no such intrinsic properties have

been discovered5.

The most readily apparent intrinsic properties of organisms are gross physical

features. Morphological species concepts classify organisms according to physical trait

similarity and functions as the everyday notion of identifying species (e.g. looking at

them). Coloration, body shape, and number of limbs are all morphological properties of

organisms. Classifying species according to morphology is, however, extremely

problematic (Ridley 2004, 348-50).

Some species contain morphologically distinct organisms. Besides the obvious

physical differences between males and females, members of one butterfly species may

mimic several distinct appearances so that they look radically different from each other,

but much like other species (Sterelny and Griffiths 1999, 184). Sibling species are

species that are nearly morphologically identical but do not interbreed, and so are

considered separate species (Okasha 2002, 196; Dupré 1981, 83). Sibling species are

problematic because they appear much more similar to each other than members of

morphologically diverse species do. Lastly, morphology does not take into account the

evolutionary history of morphologically similar species. Two populations with radically

different evolutionary paths may, by coincidence, happen to appear similar, but they

ought to be taxonomically distinguished (Lakoff 1987, 185-95). Morphological similarity

proves a weak candidate for the monistic species relation.

5 Some, such as Paul E. Griffiths (1999) and T.E. Wilkerson (1995) and Michael Devitt (2008), do remain optimistic about an essentialist program.

19

Genetic criteria for species membership were once appealing to philosophers

because genes are in part responsible for physical traits and get passed on to offspring.

Thus genes were seen to partially account for morphology while being sensitive to

descent. Unfortunately for this proposal, the maturation of genetic sciences revealed the

many genetic differences within species that undermines classification into kinds (Okasha

2002, 197). As one devoted believer in the genetic theory of species admits, there may be

as many genetic species as there are individual organisms (Wilkerson 1995, 133). The

high frequency of genetic variation within species has prompted most philosophers to

abandon the hope that intrinsic or microstructural properties can be appropriately utilized

for typological species concepts, however. Especially given that a proposal such as

Wilkerson’s neither identifies robust kinds, species with many members, nor grounds

predictions about classes of entities based on the observation of relatively few samples.

The problem for intrinsic properties as classificatory properties is that they simply

do not seem to identify species taxa. Philosophers have increasingly been motivated to

consider relational or extrinsic properties as basis for classification (Griffiths 1999,

Okasha 2002). Roughly speaking, intrinsic properties are properties an object possesses

on its own, independent of any other circumstance. Possession of relational properties, on

the other hand, might depend on how some other entity is (Lewis 1983, 111-12). I, for

example, have two legs simply because I posses them, whereas I am an uncle because my

siblings have children. Being bipedal is intrinsic to me because only an alteration to my

body could change that property, whereas my being an uncle depends on my relation to

others. It is hypothetically possible that I not be an uncle without any change in my

intrinsic properties, should my siblings never had children. Familial relations are

20

paradigm examples of relational properties that ground some species concepts. Another is

the ability to produced fertile offspring with other organisms.

The biological species concept defines species by way of breeding relations. Any

organism is said to be of a species if it can breed with other members of that species so

that differing species, which by definition cannot breed, are said to be reproductively

isolated (Ridley 2004, 351; Mayr 1992, 17). The biological species concept is riddled

with problems. As is often noted, asexual organisms simply do not breed, so it cannot

apply to those species. Either asexual do not form species, since they have no breeding

relations at all, or the BSC is woefully inadequate by leaving many taxa unrecognized

(Dupré 1993, 46). Secondly, hybridization, the process of two distinct species being

crossbred to produce some other species, is conceptually impossible according to this

account because different species are not supposed to breed (Dupré 1981, 86). Finally,

some species are reproductively compatible with some but not all of the members of their

species. In some cases animal A may reproduce with animal B, and B with some further

animal C, and yet A and C will be unable to breed. These are called ring species (Sterelny

and Griffiths 1999, 189; Ridley 2004, 372).

While the biological species concept is deficient as an account of what property

each member of a species should posses it is part of a revolution in biological thinking

about species. Rather than treating species as groups of organisms sharing some property,

they might be construed as relationally defined entities (Okasha 2002, 191). The

popularity of construing species as individuals is a result of just this insight. Mereology

takes parts of objects to compose an individual when they stand in the proper relation to

each other rather than having some feature in common. My heart is a part of my body

21

because it is embedded in a system with important functional relations to other body

parts, not because every part of my body has anything in common (though this may also

be the case).

Historical and Individualistic Monism

The shift in focus from intrinsic to relational properties has propagated many new

species theories that focus on the historical aspects of species. Species are thought to be

lineages of related organisms (de Queiroz 1999; Griffiths 1999). Historical and

individualistic accounts of species are not by definition monistic, but some believe that

some from of monism is made available by these theories (Ghiselin 1987; Hull 1976).

Individualism is not itself a full-blown theory of species since it only says that

species members (or parts) are defined by their mereological relations. Without some

further explication of what specific relation is definitional for species, individualism is a

hopelessly vague theory (Slater ms. 6). Evolutionary theory says that all of the organisms

alive today are the products of some early ancestor in remote biological antiquity, so all

organisms are genealogically related and historically connected to some degree (Dupré

1993, 48; Brogaard 2004, 231). What degree of relation or historical connectedness is

sufficient for species membership? One step toward solving this problem is by specifying

that more inclusive taxa belong to the rank of family and genera rather than species. A

ranking criterion that distinguishes species from higher categories does alleviate some

ambiguity, but it does not necessarily result in monism.

In order to reflect the diversity of causal agents directing evolutionary differentiation in different lineages, no universal ranking criterion can be found…The currently favoured monistic ranking concept of absolute reproductive

22

isolation is not the most appropriate for all groups of organisms (Mishler and Brandon 1987, 308-9).

It should also be noted that the rank problem extends to less inclusive populations—it is

not clear how small and exclusive species should be. “It is not entirely clear how we

should motivate any stopping point in constructing genealogies until we reach the

individual organism” (Dupré 1993, 49).

Another strategy for narrowing historical species concepts is to identify those

evolutionary forces that give a species cohesion and similarity. One group of organisms

may qualify as a species because in addition to sharing a common ancestor, as many

overlapping groups of organisms do, are reproductively compatible. In fact, many species

concepts can be conjoined with historical or individualist principles (Mishler and

Donoghue 1982; Mishler and Brandon 1987). Marc Ereshefsky argues for a form of

pluralism that recognizes ancestor-sharing taxa formed by different evolutionary

mechanisms (Ereshefsky, 1992). On his theory, the ecological species concept—which

classifies based on how organisms relate to their environment—or the biological species

concept can be combined with a historical model of species to identify taxa produced by

different causes. Similarly, a species individualist might hold that different individuals

are made whole by different mechanisms (Hull 1976, 178-180; Mishler and Brandon

1987; Brogaard 2004, 236-240). Because historical and individualist accounts of species

often employ any number of other species theories to supplement them, they suffer many

of the same pluralistic problems as typological accounts. There are many accounts of

species on the table that classify the same organisms in different ways whether one

23

accepts that species must form historical lineages, compose unified individuals, or kinds.

Each theory still seems to capture only part of what it means to be a species.

Finally, historical theories do not always allow for an objective demarcation of

species that matches commonly held beliefs about what the taxa should be. A surprising

example of higher taxa is illustrative: Crocodiles, as it turns out, are more closely related

to birds than any other reptile (Sterelny and Griffiths 1999, 198; Lakoff 1987, 186).

There is a common ancestor that birds and crocodiles share which no modern reptile

does. If the historical and genealogical properties of taxa are to determine membership,

no taxa can unite crocodiles with other reptiles to the exclusion of birds. Any attempt to

recognize such a paraphyletic taxa—one that does not include every member of a group’s

common ancestor—will be ad hoc. The principles supporting a monistic species theory,

however, need to apply to each and every species taxa. Given that birds are not properly

considered reptiles, which would unite crocodiles with their kin, and crocodiles are not

birds, as the history would seem to suggest, historical taxa require substantial supplement

in order to pick out those groups usually considered legitimate biological taxa. Similar

difficulties will occur for species when, at the population level, there are closely related

groups of organisms differentiated by morphological, genetic, and behavioral properties

such that they ought not to qualify as members of the same species.

Kitcher takes another strategy for rejecting historical accounts of species as

plausibly monistic by arguing that in some cases historically disconnected organisms

form species. By giving a convincing counterexample, Kitcher thinks that he can

demonstrate that even though genealogical relations are important and perhaps even

necessary for some species taxa, others will need recourse to a typological theory. Thus

24

species taxa are necessarily pluralistic because some are determined by historical

relations while others are not. Consider a hypothetical involving C. Tesselatus, a lizard

hybrid produced by two other lizard species:

Imagine that the entire initial population of C. Tessalatus was wiped out and that the species was rederived after a second incident of hybridization… Supposing that the clones founded in the first hybridization fall within the same range of genetic (morphological, behavioral, ecological) variation present in the population that has persisted to the present (Kitcher 1984, 315). There is no reason, he goes on, to consider these two hybrid populations different

species just because they do not form a continuous lineage. We might take the

hypothetical further to drive this point home. Suppose that C. Tessalatus is generated

regularly by hybridization but that each population dies out before the next comes into

existence. Suppose also, to strengthen the argument, that each population is nearly

identical to the others so that far less genetic, ecological, or morphological variability is

found in successive generations of C. Tessalatus than any other species ever observed.

Here it seems just shy of ridiculous to believe that each generation of lizards ought to be

considered a different species.

I have not, of course, entertained or rejected every theory of species monism.

However, substantial and intractable difficulties remain for the general taxonomic

approaches of typological, historical, and mereological theories of species monism.

Furthermore, specific accounts of these general approaches are unable to identify all of

and only the legitimate species taxa.

For Pluralism

25

Negative arguments against monistic accounts of species are important in terms of

the historical dispute between monists and pluralists because each blow to monism

strengthens the relative plausibility of pluralism. Some philosophers will inevitably

contend that there is still a hope of discovering some satisfactory account of species

monism, but hopes and mere possibilities brook no argument. A tenable account of

species monism should meet some minimal burden of proof. Pluralism should have its

own basis of legitimacy as well, so positive arguments for pluralism must be considered.

The most common line of reasoning offered is that no one species concept is legitimate in

any important way that other theories are not.

Many species concepts seem equally legitimate or, conversely, equally flawed.

The biological and ecological species concepts both identify factors that contribute to the

formation of species, which seems to be why both identify at least some species taxa, but

neither identify all of them. In addition, scientists employ many different species theories

depending on their needs and contexts. The validity of these modes of scientific

investigation and the predictions they underwrite seem to indicate the validity of those

many species concepts (Kitcher 1984, 326). Studying the fossilized remains of long

extinct species does not lend itself to determining the reproductive relations of those

organisms, yet such research is the predominant basis of theorizing about those species

taxa. If paleontologists can be said to discover species taxa, it would seem to bolster the

legitimacy of the morphological species concept. The monist might counter that

morphology is an indicator of species taxa but not determinant of them. However, each

case of sibling species that are completely reproductively isolated yet physiologically

similar is a strike against this line of reasoning. Interbreeding plants of drastically

26

dissimilar physiologies are similarly troublesome since their morphology does not

indicate their reproductive compatibilities or historical relations. “In some groups there is

complete reproductive isolation between populations that would be recognized as one

species on morphological grounds” (Mishler and Donoghue 1982, 494).

The utility of successful scientific practice and prediction may indicate the

metaphysical legitimacy of the species theories they presuppose, and biologists have no

small reserve of handy species theories. Thus the scientific practicability of some

theories, for Kitcher, supports the conclusion that the taxa they posit are real. His

formulation of the argument rejects monism on the basis that no species concept can

serve the purposes of all biologists, and this is based on the assumption that the plurality

of successful scientific practices implies a plurality of legitimate species concepts. But

why stop at the practical virtues of scientific taxa? Surely the ability to predict and

generalize about organisms based on scientific taxonomy lends some credence to those

taxa, but mightn’t other human activities presuppose species taxa and ground

generalizations?

Dupré argues for the validity of some taxonomic groups that biologists reject.

These taxa are taken to be legitimate because they underwrite human activities and

highlight objective similarities and differences in organisms. Lilies are an especially

complicated occurrence of a biologically disjoint class:

the Lonely Lily belongs to the genus Eremocrinum, the Avalanche Lily to the genus Erythronium, the Adobe Lily to the genus Fritillaria, and the Desert Lilly to the genus Hesperocallis. The White and Yellow Globe Lilies and the Sego Lily belong to the genus Calochortus; but this genus is shared with various species of Mariposa Tulip (Dupré 1981, 74).

27

There is a higher taxon, the family, to which all lilies belong, but it happens to include

onions and garlic. To count them as lilies would stretch the meaning of the term lily

beyond all significance. Like reptiles, lilies are a disjoint type of organism that defies

commonly accepted norms of classification. Biologists are content to say that there just is

no taxa that all and only the lilies belong to, but Dupré thinks that the concept is too

useful to do away with. Also like reptiles, the taxa we call lilies seems a legitimate and

useful one despite theoretical difficulties. Where they differ is that lily plays no crucial

scientific role.

The practical significance of the distinction between onions and garlic is lost on

the biological sciences, but the practice of food preparation would be severely hampered

without those taxonomic categories. Dupré thus argues that the practicability of common

sense taxa supports what he calls promiscuous realism, a radical variety of pluralism that

endorses a wide plurality biological taxa (Dupré 1981; 1993; 1996). Kitcher’s theory of

species, by virtue of focusing only on scientifically useful species concepts, posits fewer

overlapping taxa at the cost of ignoring common taxonomic distinctions, while Dupré’s

theory embraces a host of unscientific taxa.

A stronger appeal to the equal legitimacy of multiple species theories is grounded

in the plurality of evolutionary causes. One thing that distinguishes some scientific

species concepts from other theories is that they highlight evolutionary mechanisms. The

flavors and digestibility of certain species, while important for cooking, are not important

features of the evolutionary process. They do not tend to affect the evolution of groups of

organisms. Gene flow, breeding relations, and ecological pressures are some of the

primary actors in the process that produces species, however.

28

Connecting taxa with their causes is in principle a legitimate strategy. Species are

differentiated by their possession of different properties, and these differences are cause

by evolutionary mechanisms. When some mechanism, say environmental pressure E,

causes a substantial change in part of a population so that there are now two noticeably

distinct groups, then E can be said to have caused the formation of a new species.

Imagine a simplified universe where only one causal process actively produced variation

in the biological world. If environmental pressures somehow accounted for the entire

evolutionary process, we could identify every species ever produced by identifying only

the instances of environmentally caused variation.

If, as it seems, linking evolutionary causes to species concepts is one way of

identifying real taxa, then pluralism is sure to follow. Because no biologist would dispute

the claim that there are many evolutionary causes, we can conclude that there is a

plurality of species taxa. Furthermore, since particular evolutionary mechanisms do not

operate in isolation from other causes of variation, there must be multiple overlapping

schemes of species delineation. Some species will have genetic variations influenced by

environmental pressures, for instance, and these species can in principle be classified

based on either genetic or ecological properties, but there is no guarantee that the groups

they identify will be identical. Some individuals may share genes without sharing the

ecological niche or vice versa. Finally, there may be more evolutionary causes than just

those commonly presented in the biological sciences. Even the property of flavor may

come into play given that particularly tasty organisms are likely to have an unfortunate

disposition to be eaten often. While there is always the possibility that the various

29

interactive causes of speciation events happen to produce a monistic taxonomy, it is not

likely.

Pluralists contend that there is an overlapping plurality of species by arguing for

the legitimacy of multiple concepts. For a realist about species, to say that these theories

are equally legitimate is just another way of saying that the taxa they posit are equally

real. The equal legitimacy strategy identifies the reasons philosophers feel justified in

asserting the reality of any species taxa and goes on to claim that many species concepts

meet those standards of reality. The three appeals to equal legitimacy I examine

correspond to the three principles of moderation discussed in the following chapter. I

argue for the weakness of these positions as moderating principles, but that weakness

offers no comfort to the monist opposing the equal legitimacy arguments, as my criticism

entails that the three principles are not pluralistic enough.

Gaining Perspective on Evolution

Another less utilized argumentative strategy for pluralists appeals to the “facts on

the ground” in evolutionary biology. Some monists have insisted that species pluralism

only appeals to philosophers who have failed to familiarize themselves with the

biological facts.

“If one is lazy, one may not feel inclined to do the work. If one is incompetent, one may not succeed. If one is dishonest, one may not wish to admit that one does not know all the answers. Therefore a species is, by definition, whatever some expert finds it expedient to label with a specific epithet… The experts will disagree, not because they differ upon a scientific issue, but because they are not engaged in science at all (Ghiselin 1987, 135-6). Pluralists are, however, able to make the same point against dogmatic monists.

Many biological facts support pluralism, and some of them have been indicated already.

30

Lazyness and dishonest ignorance is just as useful for hopeful monists unable or

unwilling to come to terms with pluralism. For some philosophers, when we stand back

and take a broad view of evolution, the biological facts clearly indicate the plurality of

species.

Dupré characterizes biological ontology as a disorderly process that results in

disorderly organization. “Thus my motivation for pluralism is not methodological… but

ontological. It is that the complexity and variety of the biological world is such that only

a pluralistic approach is likely to prove adequate for its investigation” (Dupré 1993, 53).

Consider the tree of life, a graphic representation of species as branches in a family tree

spanning the entire history of life on Earth. The lines that represent each species branch

further away when the species diverge in their respective appearances and properties.

Because individual organisms within a species vary, we can picture a scattering of dots

around species each branch to represent the individuals. Taking away the lines that

represent general similarities and difference in species we are presented with a disorderly

array of dots (organisms). Some dots will form a recognizable branch of species taxa.

There are few cases of mammals that look much like humans, so the human branch will

be a tight cluster of dots. Other taxa will be more problematic. In some cases, we are

likely to see a chaotic jumble of individuals that do not outline distinct branches at all.

Diversity within species and similarity between them is profligate, and as a result there

will be borderline or indeterminate cases of species membership (Dupré 1981, 89-90).

Thus, different theories will legitimately identify different taxa.

Mishler and Donoghue consider the distribution of properties salient to species

classification, arguing that none are plausibly considered primary to the others (Mishler

31

and Donoghue 1982, 495). What motivates their pluralism is the noncorrespondence of

these properties in species taxa. As has been pointed out, organisms similar in one respect

thought important for classification (ecology, reproductive compatibility, morphology,

etc.) are often dissimilar in others.

Finally, recalling the plurality of evolutionary causes and their interactive nature,

anyone concerned with the species controversy ought to recognize that no single cause is

ever responsible for an evolutionary divergence. When some new trait arises among a

population it will have both a genetic and non-genetic factors that figure into its evolution

(Sterelny and Giffiths 1999, 98). The genetic causes of a trait are not independent of

environment in which it occurs; genes may be affected by both other genetic factors, its

genetic environment, and the ecological environment in which the organism finds itself.

If we want to understand the causes of the differences in shape, size, color, behavior, and physiology among individuals we must be prepared to work with genetic differences at many gene loci, each of small effect, and with interactions between gene and environment (Lewontin 2000, 120). Given the nature of the evolutionary process, identifying the causes of diversity in

the biological world will not necessarily identify their products, and only a pluralistic

approach is suited to capture the complexity and diversity of kinds of organisms. Mishler

and Donoghue are so committed to the principle of diversity of species taxa that they

recommend scientists “working on relatively little known organisms should not assume

that [species] concepts derived from other groups of organisms are necessarily

applicable” (1982, 501). They recommend a kind of open-ended pluralism that always

leaves room for an additional theory of taxa delineation.

32

The ongoing dispute between monists and pluralists is characterized by an

incongruity of argumentative strategies. Where monists are concerned to put forth some

theory that can account for all species taxa and argue against the general coherency of

pluralism, pluralists must show the inadequacy of particular varieties of monism and

contend that pluralism is not only consistent with but also necessitated by the biological

and evolutionary facts.

33

CHAPTER THREE

The Problem of Moderation

An important feature of an account of species pluralism is how many species taxa

it is willing to endorse. Once pluralism about species is espoused, the challenge is to find

a principled way to restrict which species concepts are to count as genuine species taxa.

Moderate theories will posit only a few real species taxa while radical pluralists believe

there are many. Unless every possible species concept identifies real taxa, pluralism must

be restricted such that only a limited number of species concepts qualify as legitimate.

Thus an important challenge facing species pluralists is to find a principle of

moderation— some criteria that species concepts must meet in order to pick out real taxa.

Some monists insist that any form of species pluralism is incompatible with

realism (Ereshefsky 2007, 13; Hull 1990, 84). The concern is that once philosophers

reject monism they reject the only viable means of arguing for the exclusion of other

species theories. Instead of insisting that just one species concept is real, the pluralist will

need a principle of moderation that identifies which taxa are real and which are not.

Devoted monists contend that whatever principle of moderation pluralists adopt, it will

ultimately constitute an arbitrary or merely subjective choice and thus will be inconsistent

with realism.

In effect this means that one can pick and choose among a variety of criteria, such as reproductive isolation, and similarities and differences in this, that, and the other. But we are not told how to make the criterion of membership be an objective one. Such pluralism does not characterize such terms as “atom” and “molecule” (Ghiselin 1987, 136).

34

I concur. But not, as Ghiselin would prefer, because species have to be monistic or

because all forms of pluralism are irreconcilable with realism. To the contrary, moderate

theories of species pluralism are not pluralistic enough.

Construing species theories, monistic or pluralistic, as claims about sets of

organisms enables every possible species theory to be evaluated (Kitcher 1987, 187). Sets

make no claim about the relationships between members. Kind theories assume that

members must share some common feature and individualist concepts rely on the

parthood relation, but sets can be any collection of organisms whatsoever. Thus set

theory provides an ontologically neutral vantage point from which to evaluate any species

theory. If species are individuals they can be represented as sets of organisms that share

the parthood relation, and if they are kinds species are construed as sets of organisms that

share a common property. Furthermore, the bare claim that species are sets recognizes

that any collection organisms might form species taxa and therefore represents the most

radically pluralistic theory possible. It is, therefore, a natural starting place for inquiring

about moderation.

Monists and pluralists alike are mortified by the claim that any set of organisms

constitutes a species taxon (Kitcher 1987, 187; Ereshefsky 1992, 358). This thesis, which

I call species universalism, conflicts with our common sense intuition that only non-

trivially similar organisms ought to be classified together. As I have noted previously,

any set of organisms are similar in some way. All organisms are descended from common

ancestry, for example, so this similarity is trivial for classificatory purposes. Nonetheless,

species universalism holds that taxa united by trivial species relations are real, if

cumbersome. Furthermore, mere sets do not conform to the intuitions that species are the

35

main unit of evolution studied by scientists. Mere sets do not evolve and they are not the

phenomena scientists are interested to investigate when they inquire about species6. The

intuitions that undermine species universalism correspond to the motivations behind the

most prominent principles of moderation found within the pluralist literature.

Three forms of moderate pluralism, defined by their respective principles of

moderation, dominate the landscape. Pragmatic pluralism holds that only sets of

organisms that are similar in interesting or useful ways form species taxa. Dupré’s

promiscuous realism is one form of pragmatic pluralism and, as the name implies, it is a

relatively radical variety of species pluralism, for it endorses more sets of organisms as

real species taxa than most philosophers are willing to ascent to. While I do not endorse

monism, I argue that monists correctly assert that no principle of moderation can both

restrict species pluralism and be consistent with realism. Pragmatic pluralism, which

claims that human interests determine which sets of organisms are species, is strait-

forwardly in conflict with realism. Interests are located in the mind of persons, not the

objective world, and for this reason monists have not been unfair in criticizing some

forms of pluralism as subjectivist. Likewise, the usefulness of a species concept is based

on how people are disposed to regard it, a feature inapplicable to metaphysical reality.

Another way to restrict species theories is to rely on the sciences for taxa

delineation. One reason is epistemological— we might think that the sciences are in the

best position to decide which species taxa are real. Even if current species concepts do

not mirror the exact structure of real species taxa, they may still provide the best

6 I say “mere sets” because both Kitcher and I contend that sets of organisms are species, hence the scientifically interesting things that evolve are sets. But only a subset of all the possible sets constitute those groups (Kitcher 1984, 309).

36

understanding of real species. Thus current species theories may be instrumental in

indicating real species taxa (Kitcher 1984, 318). It is not altogether clear which sets of

organisms should qualify as real on Kitcher’s theory. Perhaps “the most accurate

definition of ‘species’ is the cynic’s. Species are those groups of organisms which are

recognized as species by competent taxonomists” (Ibid. 308). Widespread disagreement

among scientists has, however, been a persistent problem in the species debate, so it’s not

clear that the cynic’s definition is constructive. It may be that scientists someday arrive at

a complete understanding of species so that only those sets they identify at that time

should be counted real. Without strong evidentiary support, this claim constitutes a mere

hope, much like the faith some dogmatic monists put in the future triumph of some single

species concept.

The other argument that aims to establish the claim that scientific species

concepts are real while non-scientific taxa are not involves an appeal to causality.

Scientific study of species often focuses on the causal processes that result in similarity

and diversity in the biological world. One reason for thinking that scientific species

concepts are metaphysically privileged, or more likely to be real, is that they highlight

these causal processes (Ereshefsky 1992, 359). I reject the casual approach because no

comprehensive account of every causal mechanism responsible for species formation is

offered. If some causal mechanisms are worthy of species concepts that denote real taxa,

then all of them are. However, no causal approach to species pluralism can identify every

evolutionary cause of species taxa and, therefore, constitute a properly restricted

pluralistic theory.

37

For each approach to moderation—pragmatic, epistemological, and

causal—I offer general arguments in opposition as well as a review of popular examples

of each. Dupré is a strait-forward pragmatist and I critique his theory of promiscuous

realism as such. Ereshefsky presents a causal appeal to scientific taxa, which I save for

last. Kitcher’s position is harder to pin down. I present two alternative readings of his

brand of pluralism. First I consider the pragmatic aspect of Kitcher’s moderate pluralism,

then the epistemological reading that appeals to scientific practice for moderation.

Pragmatic Pluralism

As Kyle Stanford has argued, pragmatic pluralism of the sort offered by

philosophers such as Dupré and Kitcher conflict with realism about species taxa

(Stanford 1995, 77-86). These theories rely on the idea that some sets of organisms are

not interesting or useful enough to be considered real taxa. While pragmatism describes

the way people ought choose between species concepts, choosing those that are useful for

their purposes, it is not an acceptable form of metaphysical realism about the ontology of

species.

Phillip Kitcher proposes a moderate form of pragmatic pluralistic realism by

defending four claims: (1) Species are sets. (2) There are many real biologically

interesting relations among organisms that can be used to classify species. (3) A

pluralism free from inevitable confusion and arbitrary choice is made possible by (2). (4)

This form pluralism is compatible with realism (Kitcher 1984, 309). The first point is not

of concern to this discussion, but claims (2) through (4) are problematic. (3) and (4) claim

that (2) allows for a principled restriction on which theories are to count as real species.

38

But, as Stanford points out, it is the phrase “biologically interesting” that does the work

of moderating this form of pluralism, and there are many real biologically uninteresting

relations among organisms that can be used to classify species as well (Stanford 1995,

77). (2) is in conflict with (4) because being interesting is not a feature of mind-

independent reality.

Though Kitcher rightly claims that “different views of species may be produced

by different biological priorities,” different views are also produced by decidedly non-

biological priorities as well (Kitcher 1984, p. 324).

There is no reason for privileging some set of biological categories over well-grounded common-sense categories… When these categories are taken seriously they nicely illustrate the reasons for doubting the existence of a privileged set of scientific categories, and contribute significantly to the elaboration and illustration of promiscuous realism (Dupré 1996, 444).

Dupré has advocated a form of pluralism that acknowledges all human interests as

legitimate so that even the culinary properties of organisms, properties relevant to

cooking, can be used as classificatory properties (Dupré 1981, p. 80). While the relations

that give rise to culinary properties are real, Kitcher wants to restrict his theory to include

only those properties that are interesting to biological sciences. While Dupré’s pluralism

is decidedly more radical than Kitcher’s, they both rely on the idea that some properties

are interesting or useful to some domain of human inquiry, and that these properties are

the only ones that ought to be used for classificatory practices. I call this kind of theory

pragmatic pluralism because it is moderated by reference to which properties are useful

to human activities, and I contend that it is at odds with realism.

39

Species concepts may be interesting to some and not others7. Culinary properties

are interesting to chefs, not biologists, but we may just as well think of cases where a

species concept is interesting to some biologists and not others. Historically, biologists

have changed their minds about which concepts are useful, but it cannot be the case that

these discarded theories once delineated real taxa but have since ceased to do so. Species

simply do not bounce in and out of existence based on human classificatory practices.

Such a “realism” depends on how people regard organisms, not how they are in mind-

independent reality (Stanford 1995, 83). Pragmatism does not underwrite realism because

it turns on properties that are not mind-independent. The property of being interesting is

one that only exists if some person is interested, so the fact that some features of

organisms are uninteresting says something about human mental dispositions, not the

organisms themselves. This kind of taxonomic approach is a strait forward violation of

realism because reality is independent of minds, thus real ontological entities cannot be

defined in terms of mind-dependent properties if they are to be considered mind-

independently real.

Kitcher’s pragmatic pluralism faces the problem of how to decide which persons’

interests are to be regarded as relevant. If the theory is a moderate one—if it does not

endorse a great number of species concepts—he will have to claim that not every feature

found interesting by someone indicates real species taxa. Whereas Dupré is willing to

admit that a radical plurality of species taxa correspond to the radical plurality of

interests, Kitcher needs a principled reason for rejecting schemes like culinary taxa as

7 Or, alternatively, to the same person at different times. A molecular biologist may find herself quite unconcerned with genetic constitution while in the kitchen, just as a chef may worry about the well adapted teeth of grizzlies while in the woods.

40

genuine species. The best possible reasons are discussed below and rejected, but as

Kitcher himself has recognized, moderate pluralism should ignore the interests of “the

inexpert, inane, and insane” (Kitcher 1987, 190). While there are objective facts about

what people find interesting, if we want to base a tenable form of species realism on

interests, what is needed are some objective facts about organisms that justify their being

considered interesting. Otherwise the interests of the inexpert, inane, and insane cannot

be ignored.

Those who employ species concepts, of course, ought to restrict which concepts

they accept and use based on their needs and interests. Biologists make use of species

taxa that highlight the similarities and differences in organisms which are relevant to their

domain of inquiry, as do chefs. Pragmatism functions well as an account of how scientists

and others choose which concepts to employ, but does not serve as a foundation for

realism. How scientists or chefs ought to choose species theories is a separate issue from

the metaphysical reality of species. Such discourse takes place on the pragmatic or

semantic level, not the metaphysical.

Marc Ereshefsky has raised two objections to Stanford’s arguments against

species realism, and though I do not intend to argue for an anti-realist thesis, the

objections apply to the critique of pragmatic realism offered above. One claim is that the

argument against pragmatic realism is too global in that it applies to classification

generally rather than species specifically. According to this line of thinking, “Stanford’s

interpretation of how classifications are chosen is doing all the work here,” not anything

about species taxa specifically (Ereshefsky 1998, 108).

41

Because the argument against pragmatic realism turns on the idea that interests

are the deciding factor in choosing classificatory schemes, Ereshefsky claims we may as

well apply this argument to all classificatory schemes, monist, pluralist, biological, or

otherwise. What the argument demands, however, is a claim about why species realism is

particularly problematic. There are two reasons why species realism is more difficult to

justify than realism about other taxonomies. One is the fact of pluralism. The fact that

there are multiple conflicting yet reasonable accounts of species underlies the problem of

moderation and classification selection. Electrons, on the other hand, cannot be

questioned on these grounds because there is no disagreement about how they ought to be

classified. Because of the dominance of a single theory in the taxonomy of physics,

monism goes unquestioned. No argument about how many real electron concepts there

are necessitates a principle of moderation that would restrict which electron concepts

ought to be considered legitimate.8 Pragmatism is invoked in response to the need for the

moderation of a pluralistic theory, a concern not shared by monists.

The second reason that the argument against pragmatic realism is species-specific

is that philosophers like Dupré and Kitcher have put forward these pragmatic positions as

species theories. Though they may reflect broader views about classification generally,

specific theories about species have been offered. If the arguments from interest

dependence that expose the problems in those species theories apply to some broader

theory of pragmatic classification, then it is a fault of the proposed pragmatic species

theories that they embrace the problematic claim that pragmatic classification is

8 There may be some dispute about whether or not electrons are real, but any example of taxanomic monism will do.

42

compatible with realism. The fact that Stanford’s argument exposes a flaw common to

any pragmatic approach to classification is nothing to count against it as an argument

against pragmatic accounts of species. Any taxonomic approach that relies on human

interests as its criterion for which taxa are legitimated is not a form of realism.

The Epistemological Appeal to Science

Kitcher’s appeal that only biologically interesting relations be used for

classification may hinge on the use of ‘biologically’ rather than ‘interesting.’ This

constitutes an appeal to the powers of science to discover real kinds rather than the

importance of human interests. The rigorous process of scientific investigation, one might

argue, ensures that scientists are in the best position to discover real species taxa. Though

this is surely the case, it is not itself an argument for the reality of taxa produced by

current biological theories.

To some extent, biologists are surely in the best epistemological position to

decide which species concepts identify real taxa because they have made a profession of

studying the kind of empirical facts relevant to the debate. However, it’s the wide

disagreement within the biological sciences that has precipitated philosophical

intervention in questions about species taxa. Biology, and related scientific fields, are

currently enveloped in disagreement about which species concepts adequately capture the

salient features of species taxa. As a result, appealing to the ability of biology to correctly

identify real species is inadequate as a moderating principle, though its worth noting that

biologists are not likely as promiscuous in their realism as Dupré. However, as long as

biologists are in a state of rampant disagreement, biology itself provides no solutions to

43

the species problem. This is precisely why philosophers have engaged in the species

debate; biology alone, as it is practiced today, seems to be inadequate for answering the

question “which species concepts identify real species taxa?”

The fact that empirical evidence alone is not enough to decide between species

concepts today does not mean that biologists are engaged in a hopeless endeavor. “If

empirical evidence does in some slight way affect our choice of species concepts, then

there is a glimmer of hope, no matter how slight, that biologists may eventually choose

the right concepts” (Ereshefsky 1998, 109). There is room for hope that biology will

eventually arrive at a consensus about which species are real. However, even if the future

emergence of such a consensus were guaranteed, it would not underwrite realism about

current biological taxonomies. Perhaps biologists will agree on exactly which species

concepts are valid, but that does not mean that any of those theories will be the ones

biologists consider important today. There may be a set of correct theories that does not

resemble any of our current taxonomic approaches. The claim that biologists are in the

best epistemological position to provide species concepts that correspond to real taxa

underwrites realism about current species taxa if and only if the biological community

can reach some consensus about which species concepts are valid and the theories

endorsed by that agreement are the ones available today. Both of these conditions are far

from certain.

Lastly, there is little reason to believe that only biological theories are legitimate

species theories because the phenomena biologists choose to study are historically

contingent and somewhat arbitrary. Genetics, for example, is a relatively young science.

There may be real genetic species, but their reality has never been dependent on the

44

existence of genetic science. If they are real now, they were real even before the advent

of genetics. Perhaps other domains of biological inquiry have yet to be discovered or are

not utilized. Consider again a taxonomic approach that utilizes properties important to

cooking practices. Such taxonomy is based on properties that can be empirically

investigated. Flavors and poisons, two features salient to culinary taxonomy, are products

of the interactions between the chemical composition of the organisms eaten and human

digestive systems. In principle, a whole branch of biology called culinary-biology could

be devoted to studying these kinds of facts. Nonetheless, the establishment of such a

science would not make it any more likely that culinary species were real.

Many features of organisms could be made the subject of systematic scientific

inquiries, but not all of them are. Without further argument as to why only the features

biologists currently find pertinent should qualify as a basis for metaphysically real taxa to

the exclusion of others, there is no reason to believe that only the taxa biologists find

interesting are real. Whether or not some features of organisms are accessible, interesting,

or important enough to be the subject of a science is not itself a reason for accepting or

rejecting those features as salient for species identification.

The Causal Appeal to Science

What makes the biological study of species more relevant to classificatory

practices than other methods of investigation may be its focus on causal mechanisms

(Ereshefsky 1992, 359). In ideal circumstances, the cause of similarity and diversity

among organisms would reveal all of and only the real species taxa. If there were only

45

one cause of the differentiation of biological kinds, a taxonomic system based on that

causal mechanism would adequately capture every relevant sense in which organisms

belonged to different species. Every similarity and difference in individuals could be

accounted for by reference to that one cause. Those ideal circumstances are not the ones

biologists are confronted with.

Ecological pressures, reproductive isolation, genetic transmission, and mutation

are just some of the forces that cause organisms to diversify and remain similar. Each of

these causes has at least one corresponding species theory as well: the ecological species

concept (Van Valen 1976), the biological species concept (Mayr, 1970), and the genetic

account of species (Wilkerson 1995, 131-36). The problem is that no one of these causal

mechanisms is responsible for every similarity and difference among organisms. Often

they are not exclusive causes of any one evolutionary change either. Ecological pressures

and genetic factors may both act, and interact, to produce a new species (Sterelny and

Griffiths 1999, 97-8).

One motivation for rejecting species monism is that no one classificatory scheme

can capture the relevance of all the causal mechanisms at work in evolution. The plurality

of evolutionary causes is problematic for species pluralism as well. Because evolutionary

mechanisms produce changes through interactions with each other, there are not distinct

taxa corresponding to each distinct mechanism. Were it the case that genes caused one set

of taxa to evolve while ecology caused another, developing separate classification

schemes for each cause would entirely account for every evolutionary event that

produced a new species. A pluralistic theory of species might then identify x number of

species concepts, where x is equal to the total number of evolutionary causes, and have

46

fully accounted for every relevant feature of species taxa. The fact is, however, that

single speciation events are produced by multiple interacting causes. As a result, there are

not several distinct taxa corresponding to the different causal mechanisms at work in

evolution. The problems for pluralism is that there is no reason to believe, given the

interaction of evolutionary causes, that providing one species concept for each cause will

account for every species taxa. Most species taxa just are not the product of a single

cause.

Taxonomic approaches that utilize evolutionary causes might be combined in

order to account for the plurality of processes that lead to species formation. Mishler and

Brandon, recognizing that evolutionary causes do not operate in isolation, propose that

their historical approach to species be supplemented with various species concepts

(Mishler and Brandon 1987, 306)9. In order to emphasize the causal processes most

prominent in the formation of species taxa, biologists should use the ecological species

concept to identify those lineages in which ecological pressures are prevalent causes of

the speciation event. Such an approach is able to highlight multiple causal mechanisms,

ecology and decent in this case, and hence may overcome the objection offered above.

While this is certainly a useful recommendation to scientists because the species

concept employed will in fact emphasize multiple relevant causes, it is difficult to

contend that these muti-cause species concepts will pick out the real taxa as well. Even if

a species concept is able to account for multiple causes, it is unlikely that one can or will

identify every factor that played some causal role in the evolutionary development of that

species. Such a theory would be cumbersome at best, impossible at worst. Secondly,

9 As discussed in chapter II “For Pluralism.”

47

ecological pressures may play a primary or secondary role in species formation. In some

cases, the role one causal mechanism plays will be prominent while in others it will be

minor. Species concepts are not dynamic in this way. Even if species concepts could

accurately represent the relative importance of various evolutionary factors, and this is

claim itself is dubious, it would be likely to necessitate a unique species concept for each

species taxa. It is likely that every species that has ever existed has come about from a

slightly different balance of evolutionary forces. Lastly, it is unclear how multi-cause

species concepts could represent either the interactive or dynamic nature of evolutionary

processes.

The prospects for moderate pluralism would be much brighter were single causal

mechanisms responsible for species taxa because an enumeration of those causes would

list all of the relevant factors for classification. Ereshefsky moderates his species

pluralism by limiting legitimate taxa to those that highlight causal mechanisms and

providing four principles of acceptability that reflect the practical demands applicable to

most scientific theories (Ereshefsky 1992, 358-361). He distinguishes between two kinds

of principles, motivating and sorting principles, which ought to restrict species pluralism.

Sorting principles are just the principles by which species are sorted or delineated, which

are given by the species concepts themselves. Motivating principles identify the causal

factor highlighted by the particular species concept invoked in the sorting principle. The

ecological species approach, for instance, sorts based the ecological niche occupied by

individuals, and is motivated by the fact that ecological pressures cause evolutionary

change as well as stability.

48

For Ereshefsky, cases of legitimate species are not limited to those concepts that

highlight causes of evolutionary change, however. Genuine taxa are merely nodes of

causal process. “Motivating principles either cite the causal processes that give rise to

lineages or the similar causally efficacious nature of those lineages” (Ibid. 359). Though

the theories explicitly endorsed all highlight causal process that form species, Ereshefsky

makes room for taxa based on the causal processes that similar organisms give rise to.

The principle that species taxa be motivated by identifying ‘causal nodes’ does little to

restrict species pluralism because it allows that any causally efficacious feature of

organisms can be utilized for classification. Culinary taxa highlight the similar causes of

organisms on the human digestive system, so even far-flung theories such as this are

motivated by the recognition of causal nodes. Culinary taxa such as onions fit the

description perfectly—they are lineages with similar causally efficacious properties.

Ereshefsky offers four restrictions on sorting and motivating principles in order to

moderate his pluralism. Motivating principles should be empirically testable, consistent

with other scientific theories generally, and “derivable from the tenets of the theory for

which the taxonomy is produced.” (Ibid. 361). Sorting principles should sort organisms

into a single taxonomy that is internally consistent so that no individual is part of multiple

species. This principle is not automatically violated by pluralism because consistency

constrains each individual species concept, not their conjunction. This is undisputed.

Likewise, few would dispute the claim that any species theory should be based on

empirically testable features and consistent with modern science. These are demands any

taxonomic theory must be held to. Not even culinary taxa should dare violate such

principles—but nothing about them undermines the legitimacy of culinary taxa. If any

49

restriction is to do the work of disqualifying species theories that are based on causally

efficacious properties but not of interest to the biological sciences, it is the derivability

principle.

What does it mean for a species theory to be derived from the scientific theory it

was produced for? Ereshefsky’s example is that species taxa based on interbreeding

relations are motivated by the causal role of interbreeding in producing stable groups of

organisms (Ibid. 359). Evolutionary theory tells us that interbreeding causes stability, so

the tenets of evolutionary theory indicate which nodes of causality are relevant to species

classification. We can derive the interbreeding species concept from evolutionary biology

because evolutionary theory says that species are produced, in part, by interbreeding

relations. This view would seem restrict species concepts to those that identify features

that cause species similarity and differentiation. Two problems arise for the derivability

principle: on the one hand, it does little to moderate pluralism, and on the other, it does

not guarantee the identification of real taxa.

The first issue is that the derivability principle requires that each legitimate

species concept be produced for some broader scientific theory. Like Kitcher’s emphasis

on biologically interesting theories, this approach dictates that only species concepts

accepted by certain branches of science are real. If a species concept is to be both derived

from and produced for a scientific theory in order for it to count as legitimate, no

unscientific species theory will do. The fact that many features of organisms, like means

of locomotion and culinary properties, could be the subject of a scientific discipline but

are not indicates the arbitrary nature of such a restriction. Some species taxa cannot be

derived from scientific theories, but only because of the contingent fact that scientists do

50

not find those properties interesting enough to be the subject of some broader scientific

theory, not because they are in principle unable to be the subject of scientific inquiry.

The second problem for the derivability principle is the observation that

individual causal processes do not form distinct species taxonomies. Identifying one

causal process that produces stability in populations of organisms does not identify one

set of species taxa produced by that process. There is good reason for suggesting that the

causal processes responsible for evolution should be used to identify species, but

isolating these mechanisms conceptually does not indicate species taxa that are

individuated metaphysically. The fact that many causal processes interact in species

formation undermines the claim that there are distinct species taxa for each cause

identified by evolutionary theory, or that classification based on these causes should

indentify taxa that are any more real than taxa that do not. Nothing about Ereshefsky’s

strategy for moderating pluralism vis-à-vis the derivability principle identifies taxa that

are privileged or more real than others. Furthermore, the only obvious reason for

classifying by evolutionary causes of species, rather than causally efficacious properties

of species, is that evolutionary causes are responsible for species formation.

The sense in which evolutionary causes are responsible for species formation is

deceptive, however, since no single cause is responsible for many individual cases of

species formation. Individual causal mechanisms like interbreeding are usually partial

causes, not fully responsible for speciation events. A causal approach to classification

would be compelling as a moderate pluralism if individual causes accounted entirely for

how some species were created, and if some limited set of cause-based species concepts

51

could account for most or all of the species recognized by biologists, but no such theory

or set of theories is available.

The species taxa identified by evolutionary causes are sorted based on features

that partially account for some species formation events, but because each individual

species theory accounts for only one distinct causal mechanism, they identify few, if any,

distinct kinds of species. The ecological approach, for example, will identify species that

were formed by genetic or interbreeding causes as well as ecological, or some interaction

between them all. What we are looking for is a complete account of what species taxa

have been produced by evolution, not what mechanisms produced them. Just because

ecological factors play some part of the causal role in producing some species does not

mean that ecological factors correctly identify a whole class of real species taxa. Again,

this would be the case only if ecological causes functioned independent of other

evolutionary factors. Evolutionary causes are not so isolated, and as a result there is no

reason to believe that species pluralism can be moderated by treating species as if they

were caused by isolated mechanisms.

While the sciences identify causal mechanisms that produce species, it is not clear

that the only species produced are those sorted according to causal evolutionary factors. It

may be the case that a wide variety of species are produced by the constant interaction of

evolutionary forces. As Dupré puts it “Nothing in evolutionary theory guarantees that

genealogy will always provide us with the distinctions we need in order to understand the

current products of evolution as opposed to the process by which they came to be”

(Dupré 1993, 51). Upon reflection, this should not come as a shock to metaphysicians

and scientific taxonomists intent on discovering objectively real taxa. After all, the

52

paradigm of objective taxonomy, elements of the periodic table, are neither identified nor

classified by the processes that produced them.

53

CHAPTER FOUR

Species Universalism and Mereology

Moderate pluralistic realism about species must face the difficult, perhaps

impossible, task of identifying a mind-independent feature of reality that distinguishes the

real species taxa from the unreal— a principle of moderation. If there is no proper way to

moderate species pluralism, no principle that picks out which species theories identify

real species kinds, then only a radical solution will do. One option is to deny that species

are real at all. I regard anti-realism as a position of last resort, though. There are other

theories of species pluralism that may save realism while conceding that moderation is

untenable. Mereological universalism has confronted the problem of moderation, for

individual objects, and produced a radically pluralistic theory consistent with realism.

Mereology is concerned with the relations between parts that compose whole

objects. A precise formulation of the inquiry is given by the special composition question

which asks what relations must hold between parts in order for them to compose a further

object (van Inwagen 1990, 21-2). As in the species debate, metaphysicians have offered

both monistic and pluralist answers to this question, and some have offered radically

pluralistic perspectives (Sider 2001, 120-132; Lewis 1986). Also like the species

question, the problem of moderation forces philosophers to ask whether pluralism about

composition can be restricted to a moderate number of theories (Markosian 2007).

Philosophers of biology have paid little attention to these pertinent developments

in metaphysics. The philosophical problems addressed by mereological universalism are

54

quite applicable to species theories. Ironically, though species have increasingly been

treated as individuals rather than kinds or classes, metaphysical discussions of parts and

wholes have been noticeably absent. Philosophers concerned with the species debate and

theories of composition are hard pressed to reconcile realism with the problematic cases

of particular entities, be they kinds or individuals. Pluralists also face the problem of

moderation, or “restriction” as it is often called in mereology. I propose a theory of

species universalism that parallels mereological universalism in order to overcome some

of the persistent difficulties concerning species. A plausible solution of the problem of

moderation has already been advanced by mereological universalists, so all that remains

is to develop a species pluralism of similar nature and see if it is a viable alternative.

Mereological and species pluralism have a few notable parallels. The problem of

moderation is sometimes stated as a problem about vagueness in mereology, and I note

that this vagueness is inherited from taking on the thesis of individualism, not due to any

specific mistake made by species individualists.10Vagueness is an issue about meaning

so, following the universalists’ lead, I turn to the semantic level fix the vagueness

problem. The semantic features of universalism alleviate problems concerning trivial

taxa, communicative confusion, and practicality. By relocating much of the species

debate at the semantic level, the pluralistic realist avoids arbitrary forms of moderation

inconsistent with realism.

Kitcher suggests a form of realism that entails the most radically plural theory of

species kinds. Though he does not defend it as a satisfying account of species because it

10 The vagueness problem for species individualists is discussed in Chapter Two “Against Monism.”

55

renders scientific species concepts metaphysically indistinguishable from mere sets of

organisms, the following “cheap” account of species realism is offered:

anyone who accepts a modest realism about sets can endorse realism about species. Organisms exist and so do sets of those organisms. The particular sets of organisms that are species exist independently of human cognition. So realism about species is trivially true.” (Kitcher 1984, 330).

The suggestion that any set of organisms is real may underwrite radically pluralistic

realism, but Kitcher’s line of thinking is that if all the sets of organisms are real, then

whichever species kinds the sciences decide are practical are real as well. Thus Kitcher

has a starting point from which the compatibility of realism and pluralism is made clear,

even if it is not his own theory. The consequence of this approach is that all the trivial

sets of organisms are real as well— all of the organisms in Texas, for example, form a

real taxa. Given that he promotes a moderate theory of species, this remark appears to be

merely tangential to Kitcher’s account. However, if no moderate theory of species

pluralism proves viable, then this cheap realism about biological sets may be of some

consolation. If the sets of organisms that are ordinarily called species turn out to have no

exclusive claim to actual existence, then we should consider adding sets to our ontology

that are not paradigm examples of species in order to make room for those that are.

Species universalism is the view that every possible set of organisms, including

paradigm cases of species, are on an equal footing with regard to reality (existence).

Species universalism has an ontological counterpart in mereological universalism, the

view that every possible collection of individuals, organic or otherwise, compose a

further whole object11.

11 The name “species universalism” is, of course, inspired by its mereological counterpart.

56

Species universalism is analogous to mereological universalism because they both

take real entities—composite objects and species respectively—to be constituted by every

possible set of lower level entities. Universalism in mereology takes every set of all

possible parts to be a real object, where possible parts are just any other object, composite

or simple. Species universalism takes every set of all the organisms in the world to be a

real species taxa. Because radically pluralistic ontologies are so rarely considered in the

biological context, it is instructive to look to mereological universalism as an analog for

species universalism. In fact, it seems to me that species universalism has been lurking in

the dark theoretical corners of mereological treatments of species for some time.

The motivation for rejecting moderate pluralism for species given in chapter three

is similar to the problems found with moderate, or restricted, theories of composition.

David Lewis argues that:

The trouble with restricted composition is as follows. It is a vague matter whether a given class satisfies our intuitive desiderata for composition. Each desideratum taken by itself is vague, and we get still more vagueness by trading them off against each other” (Lewis 1986, 212).

For all the proposed moderate criteria of composition—like cohesion, contact, and

fastening— none of them come close to identifying all and only the objects typically

considered ordinary individual objects, nor is there any moderately pluralistic

conglomerate of those theories that is fruitful (van Inwagen 1990, 56-60).12 Universalists

think that no compositional criterion or plurality of criteria can determine when

composition occurs, so there is only one conclusion: either composition always occurs or

12 Van Inwagen is not an anti-realist about composite individuals, he has his own pet theory of composition, but he critiques and rejects a substantial number of standard mereological relations.

57

it never does; either every set of individuals is a real object or none of them are (Van

Cleve 2007, 328).

Universalists reject the possibility that composition never occurs because the

implication is that no composite objects exist, and this seems wildly implausible given

that we ourselves appear to be such objects. In stead, they adopt the position that any

random scattering of material is a whole object. It’s a hard pill to swallow, but the

medicine seems to work. Universalism, while counter intuitive, is not obviously at odds

with realism, whereas all of the other proposed criteria for composition are. Mereology

resembles set theory on this view because a set is real just in case its members are real,

and universalism says that object is present just in case its parts are real13. Both

universalist theories of species and whole objects can be modeled after sets.

The motivation behind mereological universalism is often formulated as problem

about vagueness. Consider the problem of identifying the exact parts of a single cloud, a

collection of water droplets in the atmosphere (Lewis 1993, 164). Water vapor is present

throughout the entire atmosphere, including the areas proximate to the boundaries of the

cloud, so picking out the exact point at which the parts of the cloud cease and mere

diffuse water vapor begins is an exercise in futility. There is no principled way to decide

which water molecules are rightly considered parts of the cloud and which are not. Lewis

then extends this example to all individuals by pointing out that all individuals are

composed of particles and that for every object, some of those particles will be

questionable parts. Sometimes (and more often than we would like), it is vague whether

13 Of course this reverses Lewis’ universalist position in an important way as he intends to understand set theory through composition (1991).

58

or not some object is part of a further composite object. Vagueness troubles the

metaphysics of composition because the object itself cannot be vague. Thus problematic

cases of composition lead philosophers to conclude that composition always occurs in

order to avoid endorsing either vague entities or arbitrary principles of restricted

composition.

If not every class has a fusion, then we can consider two possible cases, one in which composition occurs and another in which it does not, which are connected by a ‘continuous series of cases’ …each extremely similar to the last. Since composition can never be vague, there must be a sharp cut-off in this series where composition abruptly stops occurring. But that is implausible. So composition always occurs” (Sider 2001, 122).

To compare, David Hull, a rather devoted species monist, takes mereology to be

the proper ontological theory of species and recognizes this problem of vague boundaries

for them. He responds by insisting that “comparable difficulties can be found for

organisms, and organisms are supposed to be paradigm individuals” (Hull 1976, 187). He

concludes that because such problems of vagueness are ignored in organisms, they can be

ignored for species as well. The proper reply to such an argument is that ignoring the

problem does not make it go away. Mereologists have examined the problem of

moderation and proposed a solution that reconciles realism with pluralism while

attending to the problem of vagueness. For organisms, this means that each problematic

case is in fact a composite organism, and a great many others besides. Counter-intuitions

aside, it is just as well to conclude that whatever mereologists have said about

composition goes for species. Without even knowing what the mereological theory is, an

intelligent person should be willing to blindly put their faith in the truth of their solution,

so long as it’s viable, rather than rely on ignorance. Elliot Sober comes closer to realizing

59

the universalist possibility when he admits that considering species mere sets of

organisms is difficult for him to accept because he does not think of individual objects as

mere sets, but explicitly acknowledges that perhaps he ought to (Sober 1984, 338).

I suspect that one reason philosophers have found mereology an appealing

ontological framework for species questions is that the phenomenon of composition is an

unsettled matter. Disagreement about how parts form integrated wholes reigns in the field

of mereology, and this gives the species individualists little restrictions on what kind of

mereological relations they can take to be species relations.14 Individualists recognize the

need for specific criteria of composition, but it is unclear that any solution is forthcoming.

As previously noted, Mishler and Donoghue have contended that there is an open

ended number of mechanisms by which the cohesion of an individual species might be

preserved, but that position seems to sidestep the very real problem of vagueness that

species individualism faces (Mishler and Donoghue 1982, 501). In short, claiming that

species are individuals is well nigh meaningless without a presupposed theory of

composition. A lack of consensus about how individuals are composed has allowed

species individualists to make vague claims about what species are without owning up to

the problems of that vagueness. Each problematic case of species membership is a case of

vagueness—it is vague whether or not the set including a problematic organism

composes a species. Thus vagueness in compositional theories allows problematic

species cases to go untreated by individualists. Either species are metaphysically vague, a

proposition I reject out of hand, or a better theory of species is needed.

14 Peter van Inwagen (1990), for example, considers, and rejects, a whole host of plausible principles of composition.

60

Realism

One deficiency the analogy between species and composite wholes suffers is

relative non-urgency of species compared to ordinary objects. Consider the dichotomy

the universalist presents just before declaring that composition occurs between all

existing objects: either universalism is true or there are no composite objects. Denying

universalism commits us to denying the reality of chairs, planets, and any object that

doesn’t have some other ontological theory to rely on. Humans might be individuals due

to some other feature than the composition of their parts, perhaps because we are persons

or because we are living. Nonetheless, anti-realism about such a broad group of familiar

objects is a foreboding conceptual possibility. Species taxa, on the other hand, are a

relatively narrow band of phenomena that might be thrown out more readily.

Nonetheless, we should run with a realistic account of species as far as possible before

carelessly succumbing to anti-realism.

Stanford holds that anti-realism is preferable to realism because species kinds are

subject to the interests of scientists and their aims. Since real species kinds cannot be

individuated on interest-relative grounds they must not be real (Stanford 1995, 86). But

there is an available alternative. I propose that accepting boring or uninteresting species

kinds as real would be preferable to denying realism about species. If a radically

pluralistic theory can support realism about scientifically interesting species theories by

endorsing realism about trivial species kinds as well, better to include trivial or counter-

intuitive entities in our ontology than exclude important ones.

61

The best argument for universal species realism is that it does not posit any

additional entity over and above the level of individual organisms. Species are real only

to the degree that their respective individuals are real. Evolution does not produce kinds

of organisms, not species kinds anyway, only the individuals that constitute them.

However, this differs from anti-realism because it endorses a theory of species taxa for

just what they are: collections of variously similar and dissimilar organisms. The

individual living beings are real and so are their features and properties, so recognition of

the fact that some groups of them are objectively similar and dissimilar is just to

recognize that taxa exist.

Given the choice between ontological nihilism with regard to individual objects

and mereological universalism, many sensible metaphysicians have sided with realism of

the cheap and radically pluralistic variety rather than admit that there are no real

composite wholes. If the problem of moderation is threatening to the species theory, as

Stanford argues, anti-realism is not the only alternative. Species theorists may well

retreat to universalism rather than admit that there are no real species kinds, and for the

same reasons. Some species taxa are almost undeniably real—Homo Sapiens for

example. As in mereology, once the door is opened to common sense entities the

problematic ones are sure to follow, and with them the universalist solution.

Quantifier Restriction and Vagueness

One objection that species universalism is sure to encounter is that even if species,

at the level of pure metaphysics, are merely sets of organisms, the subject has been

changed entirely. When philosophers, biologists, and laypeople talk about species they

are not talking about just any old set. They are referring to specific groups of similar

62

organisms, reproductively compatible groups, or whatever their favorite theory of

species, if they have a well-developed theory at all, specifies as species. This is

commonly called the “tower of Babel” complaint, and at it’s core is the belief that radical

pluralism results in inevitable confusion about what is being discussed when engaging in

discourse about species (Dupré 1993, 52). Talk about pigs, for instance, is clearly about

some distinct set of animals. If any set of animals qualifies as a species, then any number

of sets might correspond to the name “pigs,” and two people can hardly succeed in

understanding each other without time consuming explications of what specific set of

organisms are being called pigs.

Hull accuses the radical pluralist of obfuscating her terms without need, for if

there is one set of pigs corresponding to the ones caused by ecological pressures and

another distinct set that are capable of breeding, the pluralist ignores this distinction and

continues talking about pigs as if it were perfectly clear what she is talking about (Hull

1990, 84-5). This is no difficult problem for pluralism since linguistic distinctions can

easily be made between eco-pigs and bio-pigs.15Eliminative pluralists contend that the

term species should be done away with entirely in favor of more specific terms such as

eco-species and bio-species (Ereshefsky 1992, 357-8). Hull admits that the Babel

problem is just that “nothing is gained and much is lost by not reflecting this distinction

consistently in our terminology,” but simple revisions to the way we talk about species

can clear up any such confusion (Hull 1990, 84). Furthermore, it is monists who lose out

when pluralistic species distinctions are specified because they ignore useful distinctions

15 “bio-pig” is meant to highlight the fact that breeding relations are central to the biological species concept.

63

between non-identical taxa. Surely something is lost in ignoring the distinction between

entities like ‘eco-pigs’ and ‘bio-pigs’, even if that something is sometimes trivial.

Nothing is lost and something is gained by reflecting a plurality of species in our

terminology.

It is not commonly the case that speakers will have good reasons for explicitly

referring to specific theories of species when talking about them. Generally, we all know

what we’re talking about, whether implicitly or explicitly. Universalism relies on

quantification restriction to disambiguate a concept when needed. First, let’s get clear on

how mereological universalists formulate the linguistic aspect of universalism.

Vagueness, the problem that motivates universalism, is a semantic feature. It is

vague as to which organisms are rightly considered pigs, or at least what criterion should

pick them out. When two plausible candidate populations are linked to the word pig, it is

a problem that species theorists must confront: what pig means has become vague.

Furthermore, talk of ‘species’ is vague if there is no monist or moderately pluralistic

formulation of the meaning. It is not the case, however, that there is some things called

‘species’ or ‘pigs’ that are vague. What is vague is the language. “The only intelligible

account of vagueness locates it in our thought and language” (Lewis 1986, 212).

However, explicitly defining terms or labeling them with specifying operators like

‘eco’ and ‘bio’ is not the only way to disambiguate language. Consider students in a

philosophy class that are wary of offering an answer to the question “What do you

know?” They don’t hesitate because they don’t know anything or because they are unsure

of what they know. The resistance stems from the fact that in the philosophical context

justifying knowledge is a demanding task. Oftentimes, we take in cues from our linguistic

64

context to reach a nuanced understanding of terms (Grice 1975). One way of describing

the occurrences of context sensitive language is to say that the quantifiers of the language

have been restricted (Lewis 1986, 213). On this theory, quantifiers do a lot of linguistic

heavy lifting.

Lewis stresses that instead of restricting, or moderating, mereological pluralism,

we ought to restrict quantifiers so that the implicit meaning of “the pigs,” where “the” is

the context sensitive quantifier, specifies the domain of organisms the term pigs is

supposed to apply to. Its not clear that it matters which words do the work of restriction,

the point is that some objects a word might refer to on one occasion are ruled out by the

context of another. The tower of Babel objection assumes that people are incapable of

taking contextual meaning into account in order to be specific about which aspects of a

vague word are being invoked. People outside of the biological discourse have little

occasion to distinguish between competing species concepts or populations formed by

different evolutionary mechanisms.

Most of the time, words like ‘lily’ are just vague. Consider how many people use

the word species and how many people can give a satisfactory definition of the term.

Common ways of speaking about species may not bother with hodge-podge sets of

organisms, but they are not seriously committed to their unreality either. “At best,

[common sense] is committed to the view that the scattered objects of mereology, if they

exist, are for the most part not worth mentioning” (Rosen and Dorr 2002, 213).

Universalists overcome the apparent vagueness of pluralistic terminology by appealing to

language users’ ability to pick out meanings by intentionally, rather than explicitly,

ignoring or emphasizing aspects of meaning. Like Humpty Dumpty in Lewis Carroll’s

65

Alice’s Adventures in Wonderland, “When I use a word it means just what I choose it to

mean — neither more nor less.” Whether the other participants in the conversation know

what’s meant is another matter entirely. In an ironic turn of language, I stress that

“…accepting pluralism, even encouraging it, is not the same as promoting ambiguous

terminology.” (Hull 1990, 84). Indeed accepting any radically pluralistic ontology

demands specificity, whether implicit or explicit, in order to curb unnecessary confusion.

Again, one is struck by the coincidental promotion of ideas useful to species

universalism throughout mereological treatments of species, and monistic ones at that!

Hull seems to have no problem with saying that the naming of individuals, and hence the

naming of species, “is in general a rather arbitrary exercise. Any term can be applied to

anything one wishes” (Hull 1976, 178). In different contexts species terms will pick out

different sets depending on the interests of the speaker, and yet the taxa identified refer to

a set of beings that exist independently of our interests (Brogaard 2004, 238).

The tower of Babel objection has many responses. It is an issue that any theory

endorsing radical pluralism will have to confront, but linguistic philosophy is up to the

task. Leaning on intentionality or context to specify vague meanings is the philosophical

hard route. The objection is countered easily by insisting that words with a plurality of

commonly used meanings be explicitly labeled, as with ‘eco-pigs’ and ‘bio-pigs.’ It is

surprising that it has been raised and responded to so many times (Dupré 1993;

Ereshefsky 1992; Kitcher 1984; Hull 1990).

66

CHAPTER FIVE

Conclusion

When people talk about species they can discuss whatever organisms they care to

address. Sometimes there is no thing or taxa that corresponds to their discussion topic

because the conversation takes place at an unsophisticated level that does not necessitate

a complicated understanding of species. Other times they are referring to whatever taxa

the situation demands they consider. This is just the kind of theory some philosophers are

likely to mistake for subjectivism or nominalism. It is an easy mistake. Species

universalism supports the idea that I refer to whatever collection of organisms I want to

when talking about species, and that those taxa are real, despite their possible

ridiculousness. An immature understanding of the theory might conclude that people can

think species into existence (subjectivism), or stipulate that they are real (nominalism).

The short response is that it is just a fact that sets of organisms exist. Some sets are

reproductively compatible, some share ecological niches, one set consists of me and the

cockroach in the corner, but all of them are real (Brogaard 2004, 238).

This approach to classification attempts to transcend the debate between realism

and nominalism to take elements from each. Linguistic conventions play a role in how we

talk about and conceive of the biological world, even in the most rigorous of scientific

contexts. Hilary Putnam has made this point by saying “elements of what we call

‘language’ or ‘mind’ penetrate… deeply into what we call ‘reality’” (Putnam 1990, 28).

For the mereological universalists who thinks that there are an uncountable number of

real objects, the objects of what we might call “practical ontology” (the things we usually

67

consider real) differ from everything else only because we recognize and think about

them. The rest is completely trivial despite being real. Thus Putnam suggests that instead

of inquiring about what is, the search for metaphysical reality, philosophy ought to be

more concerned with finding better ways of thinking about reality (Ibid. 21). Reality is a

mundane feature of ontology on this view, and not worthy of devoting much

philosophical resources to. Most of the real intellectual work needs to be done at the

semantic or pragmatic level. One reason for rejecting Stanford’s appeal to anti-realism is

that he is too concerned about interest-relative ontology. All practical ontology is

interest-relative, but being interest relative is compatible with being real when so many

real taxa exist.

What distinguished scientific species concepts from the many real taxa endorsed

by radical pluralism? Scientific taxa may simply be more useful and interesting.

Scientists and laymen alike may simply ignore the mass of uninteresting taxa, much the

way uninteresting composite objects are ignored in everyday discourse according to

Lewis’ universalist theory of composition (Lewis 1986, 213). Classifying organisms

based on physical similarity may indicate the presence of other interesting properties

better than classifying based on body weight, but that does not mean morphological taxa

are any more real than size-based taxa, just more useful. All too often, the utility or

naturalness of a species theory is taken as an indicator of reality and a lack of usefulness

as metaphysically salient. Ontology is the subject of what exists, not what is useful.

Philosophers and scientists have a hard time believing that scientific taxa are not more

real than non-scientific classifications because the scientific theories are the most useful

for attaining understanding of the natural world. As well they should, it is the role of

68

natural scientists to strive for such understanding. Ontology, however, is best left to the

metaphysicians. Dupré, I suspect, has a similar view. Part of his motivation for claiming

that any useful taxonomy is real is to wrest species ontology out of the hands of science

and its dogmatic adherents (those whose believe science is in the business of providing

metaphysically privileged descriptions of reality).

In the case of the periodic table, there is no reason or use in thinking that

scientists did not discover a fundamental ontology of the universe. The classification of

substances by way of microstructural necessary and sufficient conditions fits our best

models of typological ontology. Not so for species. Biologists seem to make headway

trying to understand the mechanisms of evolutionary theory while progressing little in

their understanding of the species produced. Thankfully science has not yet co-opted

every last corner of philosophy, and for good reason. Sometimes it takes a philosopher to

point out the limitations of scientific reasoning. While biologists are perfectly capable of

comprehending what species are, their understanding of biological causes has not

produced a legitimate ontology of the evolutionary products. Due attention to the

interactive nature of evolutionary causes and their variability of importance in specific

taxa indicates that that there is no reason to expect that it would.

Philosophical recognition of the radical plurality of species taxa does not have

much, if any, practical impact. Laypersons are free to talk about species in a vague

manner so long as it suits their purposes. Scientists ought to be guarded about which

species concepts they use because it affects their ability to go about doing science. Just

because species theories that highlight evolutionary causes are not metaphysically

privileged does not mean they aren’t scientifically privileged. To the contrary, classifying

69

organisms based on their evolutionary causes is productive for doing exactly what

evolutionary biologists are supposed to do, reaching for a better understanding of the

process of evolution. While it is tempting to think that only interesting, useful, or

scientific species are real, these positions are not tenable upon philosophical reflection.

70

Bibliography

Boyd, Richard (1999) “Homeostasis, Species, and Higher Taxa” in R.A. Wilson (ed.), Species: New Interdisciplinary Essays. Cambridge: MIT Press.

Bryant, Rebecca (2000) Discovery and Decision. London: Associated University

Presses, Inc.

Brogaard, Berit (2004) “Species as Individuals”, Biology and Philosophy 19: 223-42 Chandler, Ray, and Gromko, Mark (1989) “On the Relationship Between Species

Concepts and Speciation Processes”, Systematic Zoology [38(2)]: 116-25. De Queiroz, Kevin (1999) “The General Lineage Concept of Species and the Defining

Properties of the Species Category” in R. Wilson (ed.), Species: New Interdisciplinary Essays. Cambridge: MIT Press.

Devitt, Michael (2008) “Resurrecting Biological Essentialism”, Philosophy of Science 75: 334-82.

(2009) “Natural Kinds and Biological Realisms”, forthcoming in Carving

Nature at its Joints: Topics in Contemporary Philosophy 8. Dupre, John (1981) “Natural Kinds and Biological Taxa”, The Philosophical Review 90:

66-90. (1983) “The Disunity of Science”, Mind 92: 321-46. (1993) The Disorder of Things. Cambridge: Harvard University Press. (1996) “Promiscuous Realism: Reply to Wilson”, British Journal of the

Philosophy of Science 47: 441-44. Ereshefsky, Marc (1992) “Eliminative Pluralism”, Philosophy of Science 59: 671-90. (1998) “Species Pluralism and Anti-Realism”, Philosophy of Science 65: 103-

120.

(2001) The Poverty of the Linnaean Hierarch. Cambridge: Cambridge University Press.

(2007) “Species” in Stanford Encyclopedia of Philosophy [web]. Retrieved Dec. 2008, from http://plato.stanford.edu/entries/species.

71

Ghiselin, Michael T. (1975) “A Radical Solution to the Species Problem”, Systematic

Zoology 23: 536-44 (1987) “Species Concepts, Individuality, and Objectivity”, Biology

and Philosophy 2: 127-43.

Grice, H.P. (1975) “Logic and Conversation” in Cole, Peter and Morgan, Jerry L. (eds.), Syntax and Semantics 3: 41-58. New York: Academic Press.

Griffiths, Paul E. (1999) “Squaring the Circle: Natural Kinds with Historical Essences,”

in Wilson, R.A. (ed.) Species: New Interdisciplinary Essays Cambridge: MIT Press.

Hacking, Ian (1991) “A Tradition of Natural Kinds”, Philosophical Studies 61: 109-26 Hull, David L. (1976) “Are Species Really Individuals?”, Systematic Zoology 25: 174-91. (1990) “Conceptual Selection”, Philosophical Studies 60: 77-87. Kitcher, Philip (1984) “Species”, Philosophy of Science, 51: 308-333. (1987) “Ghostly Whispers: Mayr, Ghiseli, and the “Philosophers” on the

Ontological Status of Species”, Biology and Philosophy 2: 184-92.

Kornblith, Hilary (1993) Inductive Inference and its Natural Ground. Cambridge: MIT Press.

Lakoff, George (1990) Women, Fire, and Dangerous Things. Chicago: University Of

Chicago Press. LaPorte, Joe (2003) Natural Kinds and Conceptual Change. Cambridge: Cambridge

University Press.

Lewis, David (1983) “Extrinsic Properties”, Philosophical Studies 44: 197-200. (1986) On the Plurality of Worlds. Oxford: Basil Blackwell. (1993) “Many, But Almost One” in D. Lewis (ed.), Papers in Metaphysics and

Epistemology. Cambridge University Press. Lewontin, Richard (2000) The Triple Helix: Gene, Organism, and Environment.

Cambridge: Harvard University Press, 2002.

72

Markosian, Ned (1998) “Brutal Composition”, Philosophical Studies 92: 211-49. (2007) “Restricted Composition”, Forthcoming in Hawthorne, Sider, and

Zimmerman (eds.), Contemporary Debates in Metaphysics. Basil Blackwell. Mayr, Ernst (1970) Populations, Species, and Evolution. Cambridge: Harvard University

Press.

(1987) “The Ontological Status of Species: Scientific Progress and Philosophical Terminology”, Biology and Philosophy 2: 145-66.

(1992) “Species Concepts and Their Application,” in Ereshefsky, Marc (ed.), The Units of Evolution. Cambridge: MIT Press.

Mishler, Brent D., and Brandon, Robert N. “Individuality, Pluralism, and the

Phylogenetic Species Concept,” in Hull, David L., and Ruse Michael (eds.), The Philosophy of Biology. Oxford: Oxford University Press.

Mishler, Brent D., and Donoghue, Michael J. (1982) “Species Concepts: A Case for Pluralism”, Systematic Zoology 31: 503-11.

Okasha, Samir “Darwinian Metaphysics: Species and the Question of Essentialism,” Synthese 131; pp. 191-213, 2002. Putnam, Hilary (1990) Realism with a Human Face. Cambridge: Harvard University

Press.

Richerson, Peter J., and Boyd, Robert. (2005) Not by Genes Alone: How Culture Transformed Human Evolution. Chicago: University of Chicago Press.

Ridley, Mark. (2004) Evolution (3rd ed.). Boston: Blackwell Publishing. Rosen, Gideon., and Door, Cian. (2000) “Composition as Fiction” in R. Gale (ed.), The Blackwell Guide to Metaphysics. Boston: Blackwell Publishing. Sider, Theodore. (2001) Four-Dimensionalism: An Ontology of Persistence and Time.

Oxford: Oxford University Press. Slater, Matthew H. (unpublished manuscript) "Against Species as Individuals" available

online at http://www.uidaho.edu/~mslater/research/ (unpublished manuscript) “The Arbitrariness of Mereological Immoderation”,

available online at http://www.uidaho.edu/~mslater/research/

73

Sober, Elliot (1980) “Evolution, Population Thinking, and Essentialism”, Philosophy of Science 47: 350-83.

(1984) “Sets, Species, and Evolution: Comments on Philip Kitcher’s

‘Species’”, Philosophy of Science 51: 334-341. Stanford, Kyle P. (1995) “For Pluralism and Against Realism About Species”,

Philosophy of Science 62: 70-91. Sterelny, Kim and Griffiths, Paul E. (1999) Sex and Death. Chicago: University of

Chicago Press. Van Cleve, J. (2007) “The Moon and Six-Pence: A Defense of Mereological

Universalism,” in Sider, T., Hawthorne, J., and Zimmerman, D., (eds.), Contemporary Debates in Metaphysics. Basil Blackwell.

Van Inwagen, Peter (1990) Material Being. Ithaca, N.Y.; London: Cornell University

Press. Van Valen, L. (1976) “Ecological Species, Multispecies, and Oaks” Taxon 25: 233-39. Wilkerson, T.E., (1995) Natural Kinds. Hants: Ashgate Publishing Company.